Research Projects - Radar

In numerous areas of electrical engineering such as radar, imaging and communication technology, there is currently a competition between technologies: integrated optics is competing directly with classic CMOS circuit technology for very high frequency ranges above 300 GHz. As integrated optics is not yet suitable for mass production, CMOS circuits for the so-called terahertz range are of particular interest to industry and are therefore the subject of the latest research.

Some of these circuits are to be realized as part of the DFG-funded project “TIEMPO” in cooperation with the TU Dresden and the FAU Erlangen-Nümberg. The aim is to develop a radar system in a frequency range of 220 GHz to 420 GHz. The high bandwidth makes it possible to achieve resolutions of less than one millimeter through clever signal processing.

This results in a wide range of application areas such as the recognition of gestures or material defects in products, as well as integration into the future 6G mobile communications standard, particularly with regard to self-driving cars. The development of the ultra-high frequency components in Globalfoundries' 22nm CMOS technology and the subsequent measurement of these pose particular challenges. These must be mastered in the course of the project, whereby there are numerous opportunities for final theses or student assistant positions. FPGA development, signal processing and system simulations through to millimeter wave circuit technology, antenna design and the design of so-called metasurfaces are currently the most important topics of the project. If you are interested, please contact the responsible employee (Finn Stapelfeldt).

Digital and energy-efficient radar networks for heterogenous E/E vehicle architectures

Motivation
The increasing number of networked devices and sensors, the “Internet of Things” (IoT), enables a wide range of new applications. However, it is also generating a rapidly growing volume of data. Processing data at its point of origin (edge computing) helps to handle it efficiently. Edge computing strengthens the functionality, sustainability, trustworthiness and cost-effectiveness of electronic applications through the use of artificial intelligence and networking. The aim of the OCTOPUS projects is to provide highly innovative electronics for specific applications in order to exploit these advantages.

Objectives and procedure
The aim of the project is to research the reliable detection of the vehicle environment by digital radar networks in fully automated driving. The focus is on the extensive digitization of all system-relevant functionalities and the development of an optimal computing load distribution between network controllers, sensor nodes and central computer. By using state-of-the-art FPGA (Field Programmable Gate Array) circuits, various computing steps and artificial intelligence (AI) algorithms are to be implemented in signal processing. The computing load distribution is to be optimized with regard to image quality, costs, energy efficiency, reliability and real-time capability. A demonstration of a smart radar sensor network is planned at the end of the project.

Innovations and prospects
The project makes an important contribution to the development of sustainable communication technology and thus supports the achievement of climate protection goals. First and foremost, it enables the resource-efficient and safe use of communication technology for fully automated driving. However, the project results are also transferable to other sensor types and applications, such as Industry 4.0, logistics and medical technology.
If you are interested, please contact the responsible employee (Max Leander Prause).

Sensory occupant detection for safe autonomous driving using radar

The introduction of safe autonomous vehicles requires reliable occupant monitoring (“in-cabin sensing”). Radar sensors are very well suited for this and have clear acceptance advantages over camera-based systems due to their inherent privacy protection. However, the spatial, speed and angular resolution of the sensors must be significantly improved compared to the state of the art. This requires an increase in usable bandwidth, operating frequency and the number of virtual reception channels.

In SICHER, a high-resolution radar sensor with a highly integrated broadband transceiver in SiGe BICMOS technology will therefore be developed as a key component. In addition to technology and circuit development, the development of AI-based algorithms for object recognition and classification as well as for the detection of gestures and vital parameters for direct attention assessment is a focal point. They are to be implemented close to the sensor in embedded AI accelerators.

The performance of the multifunctional sensor technology, including the AI algorithms for deriving vital parameters, is demonstrated in the laboratory and in the test vehicle.

SICHER is one of the four projects in the “NovoMotive” family, which has set itself the goal of using standards to create technology standards for tomorrow's automotive electronics systems. The use of synergies through the exchange and transfer of results within the project family increases the benefits of the project results and enables a sustainable influence on the entire automotive industry.